Spatial and Seasonal Response of the Surface Area-to ... · Rabat-Instituts, 10000 Rabat, Morocco. 4 . Département de Statistique et Informatique Appliquées, Institut Agronomique

International Journal of Applied Engineering Research ISSN 0973-4562 Volume 12, Number 21 (2017) pp. 10676-10692

© Research India Publications. http://www.ripublication.com

10676

Spatial and Seasonal Response of the Surface Area-to-Volume Ratio to

Changes in Moisture Content in Some Dominant Mediterranean Forest

Fuels

Salaheddine Essaghi1,2,*, a, M’hamed Hachmi2,b, Mohammed Yessef 3,c, Mohammed Dehhaoui4,d and Abdessadek Sesbou2,e

1 U.R. Gestion Conservatoire des Eaux et des Sols, Institut Agronomique et Vétérinaire Hassan II BP 6202, Rabat-Instituts, 10000 Rabat, Morocco.

2 Département du Développement Forestier, Ecole Nationale Forestière d’Ingénieurs, BP 511, Tabriquet, Salé 11015, Morocco. 3 Département des Ressources Naturelles et Environnement, Institut Agronomique et Vétérinaire Hassan II BP 6202,

Rabat-Instituts, 10000 Rabat, Morocco. 4 Département de Statistique et Informatique Appliquées, Institut Agronomique et Vétérinaire Hassan II BP 6202,

Rabat-Instituts, 10000 Rabat, Morocco. *Corresponding author

aOrcid: 0000-0001-7743-7217, bOrcid: 0000-0002-8371-0862, eOrcid: 0000-0002-9292-6745

Abstract

Surface area-to-volume ratio (SVR) strongly influences plant

flammability and it is widely used in most fire behaviour

prediction systems all over the world. In these prediction

systems, SVR was regarded only as an average of the whole

species present at the site. However, SVR is species-specific

and fluctuates strikingly according to leaf and needle moisture

content (H) depending on environmental conditions. This

situation results in inaccuracies in predicting fire behaviour and

lack of reliability of the systems used. Hence the need to model

the relation SVR vs H for each plant species, taking into

account the possible effects of the site and season. This

modelling would yield the SVR values corresponding to the

immediate H of leaves and needles, leading to a more efficient

and accurate fuel hazard assessment meeting a wide range of

H. Several leaf and needle samples were collected from thirteen

tree and shrub species over the four seasons of the year, at six

sites in western Rif Mountains, Morocco. Every season, SVR

and H were measured regularly during the drying of the

samples. SVR values were significantly affected by both site

and season for all species. Correlation between SVR and H was

significant for all species except Cistus crispus. The modelling

of the relation SVR vs H highlighted two separate groups of

species regarding the response to water stress. SVR changed

significantly under the effect of H and the environmental

conditions relating the season and site. However, SVR response

showed two distinct behaviours according to species. Further

research would extend the SVR database to other plant species,

in order to cover more ecosystems and therefore be able to

integrate it into the fire behaviour prediction systems.

Keywords: Modelling; Drying; Fire behaviour prediction

systems.

INTRODUCTION

Surface area-to-volume ratio (SVR) is considered as a

significant factor of plant flammability and a critical parameter,

required in fuel characterization and fuel hazard assessment [1–

5]. Moreover, SVR is a determining factor in heat and moisture

exchange rates [6]. This is well illustrated especially in the case

of fuels with higher SVR values, which exhibit faster water

loss. This situation reduces time to ignition and increases the

fire spread rate [4,7,8].

Fuel hazard can be assessed using the quantification of the

species pyric properties [6] which determine flammability of

the species [1,9] especially SVR.

Otherwise, SVR is a basic parameter much sought after in most

of fire behaviour prediction systems such as FARSITE [10],

BehavePlus [11], FIRETEC [12]. SVR values used in these

prediction systems are only average values of all existing

surface fuels, regardless of the ecosystem [11,13,14] and

therefore imprecise. Nevertheless, SVR varies according to the

species [1,2,4,6,15,16]. Furthermore, since SVR estimation

lean primarily on leaf and needle thickness [4], which is highly

affected by moisture content (H) [17–19], SVR is then

substantially influenced by H. The latter parameter itself

depends on fuel type, and varies during the year according to

season, climatic factors or phenological stages of the plants and

fluctuates due to hydric stress [1]. To our knowledge, no study

has so far tried to adjust SVR for the effect of H.

The improvement of fire behaviour prediction systems and the

flammability ranking of plant species, constitute two major

components of wildfires prevention and forest fire management

[20,21]. Such actions require thus an accurate SVR database for

each species taking into account the variations of H. The fuel

SVR database should also consider H differences of natural

fuels (live or dead, litter, debris etc [22]) and the seasonal

changes.

mailto:[email protected]



10677

Table 1: Distribution of the collected species among sampling sites

Sampling

sites

Ecosystem Altitude

(m)

Longitude N Latitude W Species collected

Larache (S1) Mixed stand of Pinus pinea

and Quercus suber

25 35°13’45.9’’ 6°14’25.0’’ Cistus salviifolius, Pinus pinea, Quercus suber,

Ahl Srif (S2) Mixed stand of Pinus

canariensis and Quercus suber

142 35°00’18.1’’ 5°41’26.5’’ Ceratonia siliqua, Cistus crispus, Cistus monspeliensis, Pinus canariensis, Pistacia lentiscus, Quercus suber

Souk L’Qolla

(S3)

Pure stand of Pinus pinaster 263 35°5’2.5’’ 5°34’19.5’’ Arbutus unedo, Ceratonia siliqua, Cistus albidus, Cistus monspeliensis, Pinus pinaster, Pistacia lentiscus

Tanaqoub

(S4)

Pure stand of Quercus suber 615 35°7’2.7’’ 5°26’59.1’’ Cistus monspeliensis, Quercus suber

Dardara (S5) Mixed stand of Pinus

canariensis and Quercus suber

406 35°7’50.0’’ 5°17’23.7’’ Arbutus unedo, Pinus canariensis, Pistacia lentiscus, Quercus suber

Bellota (S6) Pure stand of Pinus

canariensis 128 34°56’5.0’’ 5°31’56.1’’ Cistus monspeliensis, Pinus canariensis, Quercus

coccifera, Viburnum tinus

The purpose of this paper is to adjust SVR for the effect of H.

Testing the effects of season and site on this relationship for

each species is also aimed.

MATERIALS AND METHODS

Study sites

Six sites were identified in north-western Morocco (western Rif

Mountains). Each site had experienced no fires for at least 3

years and contained canopy and understory species

characterizing the respective ecosystems (Table 1). All the

study sites are properties managed by the Moroccan High

Commission for Forests and constitute the forests of Larache

(Site 1), Ahl Srif (Site 2), Souk L’Qolla (Site 3), Tanaqoub (Site

4), Dardara (Site 5) and Bellota forest (Site 6) (Table 1; Figure

1). Each site was chosen according to an altitudinal gradient

starting from the cork oak forests (Atlantic coast) to the pine

forest of Chefchaouen with the aim of covering most tree and

shrub species (Figure 1). The mean distance between sites was

30 km, whereas two successive sites were between 10 and 32

km apart. The farthest sites from each other were the sites 1 and

5, which were 65 km apart.

Figure 1: Map of the study area (Western Rif Mountains) in

north-western Morocco showing sites where the samples were

collected

Species selection and sampling

Canopy and understory species were chosen based on their

abundance in the ecosystems of Western Rif. The tree species

studied were Pinus pinea (stone pine), Pinus pinaster (maritime

pine), Pinus canariensis (Canary Island pine), Ceratonia siliqua (carob tree), Quercus suber (cork oak) and Quercus coccifera (kermes oak). The shrub species were Arbutus unedo

(strawberry tree), Cistus albidus (grey-leaved cistus), Cistus crispus (wrinkle-leaved rockrose), Cistus monspeliensis (narrow-leaved cistus), Cistus salviifolius (sage-leaved

rockrose), Pistacia lentiscus (mastic tree) and Viburnum tinus

(laurustinus) (Table 1).

The species were selected according to their abundance at the

site. The samples were collected in the four seasons of the year

2014 (January, April, August and November). Because leaves

and needles are considered the most flammable parts of the

plants [20], only leaves and needles have been studied in the

present work. To eliminate the possible effect of age, six leaf

and needle samples of different ages were selected from 3-4

individuals for each species at each sampling site for all

seasons. A total of 480 leaves and 120 needles was collected

from the sites, placed into sealed plastic bags and transported

in a thermally insulated box with ice.

Samples physical characteristics and H monitoring

Once at the laboratory, for all leaves and needles collected from

all sites, thickness and width at mid length and weight were

measured in six repetitions per site for each species. Thickness

was measured near the main leaf vein. Width and thickness

measurements were performed in order to calculate the

corresponding surface area-to-volume ratio (SVR) values

according to the method of Hachmi et al. [4], whereas weight

measurements would be used to calculate H-values. For each

sample and each H level, SVR was calculated using the

equation of Hachmi et al. [4]:



10678

For leaves:

𝑆𝑉𝑅 = (4 𝑡ℎ𝑖𝑐𝑘𝑛𝑒𝑠𝑠⁄ ) × (1 − 𝑒 2⁄ ) × (4 𝜋⁄ )𝐾

For needles with non-elliptic cross-section (P. canariensis):

𝑆𝑉𝑅 = (4 𝑡ℎ𝑖𝑐𝑘𝑛𝑒𝑠𝑠⁄ ) × (1 + 𝑛 𝜋⁄ )

For needles with semi-elliptic cross-section (P. pinaster and P. pinea):

𝑆𝑉𝑅 = (4 𝑡ℎ𝑖𝑐𝑘𝑛𝑒𝑠𝑠⁄ ) × ((1 − 𝑒 2⁄ ) × (4 𝜋⁄ )𝐾 + (2 𝜋⁄ ))

where e is the elongation coefficient, dimensional

characteristics of the fuel cross-section shape (𝑒 = 1 −

(𝑡ℎ𝑖𝑐𝑘𝑛𝑒𝑠𝑠 𝑤𝑖𝑑𝑡ℎ⁄ )) ;

K is a function of e; (𝐾 = [𝑒 (2 − 𝑒)]²⁄ );

n is the number of pine needles per fascicle. Both thickness and

width are expressed in meter.

The SVR calculation method of Hachmi et al. [4] provides a

direct technique to assess SVR depending on fuel type. This

method gives acceptable results in comparison with other SVR

calculation methods. In addition, its usage is easier and faster.

However, the method of Hachmi et al. [4] is less accurate than

methods requiring specialised equipment such as the image

analysis method, though the latter has the disadvantage of being

time-consuming.

To adjust SVR for the effect of H, the aforementioned

dimensions were monitored during their progressive drying.

The samples were therefore placed inside papers, slightly

pressed the first days of drying, just enough to keep their initial

shape and still adapt to dimension measurements throughout

the drying process. Dimension and weight measurements were

repeated at regular intervals until the samples got dried.

Samples were subjected to two different periods of drying:

1) Air-drying period. This period took place once the samples

arrived at the laboratory. During the first days when the drying

rate was high, the morphological and weight measurements

were taken every 12 hours. Later, as the air-drying rate

decreased considerably, the measurements were performed

every 24 hours.

2) Oven-drying period: once the samples were completely air-

dried, and in an objective to reach progressively, very low

levels of H, feigning dead foliage and litter H.

During the oven-drying period, the samples were placed in the

oven at different gradually increasing temperatures (30, 35, 40,

50 and 60°C) for 24 hours at each oven temperature. For each

oven temperature, the same measurements (thickness, width

and weight) were carried out. Weights obtained after oven-

drying samples at 60°C for 24 hours were considered as the

oven-dry weight that will establish a base of calculation of the

samples H during all drying stages.

H of each sample was computed, for all drying stages, based on

oven-dry weight [23,24].

All weights were measured at the laboratory with a Kern® ALJ

120-4 balance with a maximum weight of 120 g and 0.1 mg

accuracy. Width and thickness measurements were carried out

using a digital electronic calliper Powerfix® 0-150 mm Z11155

with a resolution of 0.01 mm, accuracy of ±0.02 mm and

repeatability of 0.01 mm.

Statistical analysis

The data collected were analysed for the effects of species,

sampling site and season using ANOVA test in IBM Statistical

Package for Social Sciences (SPSS). Pairwise comparisons of

means were performed using Scheffé’s test to separate species

indicating significant differences due to season and site effect.

Duncan’s multiple comparison test was used to check

thickness, width and SVR differences between species at each

season. Several regression models were tested to find a

correlation between SVR and leaf and needle H.

RESULTS

Physical parameters of some dominant Mediterranean

forest fuels

The average values of thickness, width and SVR for the freshly

collected samples per season are presented in Table 2. The

highest average leaf thickness was recorded for C. albidus and

V. tinus (1.27 mm) in winter, whereas the lowest leaf thickness

was found in Q. coccifera spring samples (0.15 mm). Pinus pinea needles collected in winter were the thickest (0.89 mm),

whilst P. pinaster summer needles were the thinnest (0.49 mm).

Viburnum tinus leaves collected in winter showed the greatest

average leaf width (61.14 mm), while P. pinea needles

collected in winter were the widest ones (1.60 mm). The lowest

average leaf width was found in C. monspeliensis summer

samples (4.85 mm), whereas the least wide needles were P. pinaster needles collected in spring (0.92 mm) (Table 2). One-

way ANOVA and Duncan’s multiple comparison tests revealed

significant difference between species in regard to thickness

and width. Leaf and needle thickness formed 6 levels of

significance in summer and fall, 2 levels in spring and 4 levels

in winter. Regarding width, Duncan’s comparison test showed

5 levels of significance in fall, 6 levels in winter and summer

and 8 levels in spring (Table 2).



10679

Table 2: Average values of the fresh foliar thickness, width, SVR and moisture content ± standard error (s.e) presented by

collection season for each forest fuel. SVR mean values are followed by their corresponding standard deviations (s.d.)

Test of significance was performed by the Duncan’s multiple range test, at 95% confidence level. For each season, different letters in the same column indicate significant differences among species at p < 0.05

Season Species Thickness (mm) Width (mm) SVR (m-1) s.d. Moisture content

(%)

Winter

Viburnum tinus 1.27± 0.16 d 61.14± 4.53 f 2161± 314 a 703 155± 3

Cistus albidus 1.27± 0.19 d 23.11± 2.33 c 2178± 266 a 595 126± 25

Ceratonia siliqua 1.09± 0.10 cd 29.17± 0.85 de 2508± 200 ab 633 136± 3

Cistus crispus 1.03± 0.06 bcd 11.13± 0.50 b 2550± 158 ab 388 169± 20

Quercus coccifera 0.87± 0.05 abc 32.49± 1.89 e 2997± 199 ab 444 83± 2

Cistus monspeliensis 0.91± 0.05 abc 10.06± 0.65 b 3009± 152 ab 697 193± 15

Arbutus unedo 0.88± 0.07 abc 34.00± 1.71 e 3095± 268 ab 848 132± 9

Quercus suber 0.88± 0.07 abc 27.13± 1.96 cd 3294± 246 ab 1127 125± 11

Cistus salviifolius 0.77± 0.03 ab 13.16± 0.71 b 3354± 135 ab 330 245± 32

Pistacia lentiscus 0.65± 0.05 a 11.32± 0.33 b 4260± 297 b 1150 108± 3

Pinus pinea 0.89± 0.02 abc 1.60± 0.06 a 6450± 156 c 348 176± 10

Pinus pinaster 0.71± 0.03 a 1.13± 0.02 a 8295± 285 d 638 141± 3

Pinus canariensis 0.81± 0.07 abc 1.42± 0.15 a 9997± 1062 e 4111 118± 15

Spring

Cistus crispus 0.33± 0.01 a 10.43± 1.08 bc 7681± 164 a 402 161± 12

Ceratonia siliqua 0.33± 0.01 a 37.78± 1.35 g 7801± 276 a 916 94± 4

Pinus pinea 0.67± 0.06 b 1.33± 0.06 a 8764± 828 ab 2028 151± 11

Pistacia lentiscus 0.29± 0.01 a 12.33± 0.42 c 8982± 295 ab 1216 85± 1

Pinus pinaster 0.67± 0.02 b 0.92± 0.02 a 9084± 239 ab 585 127± 4

Quercus suber 0.26± 0.01 a 27.03± 1.00 ef 10002± 288 ab 1380 46± 5

Cistus salviifolius 0.27± 0.03 a 11.26± 0.80 bc 10004± 1149 ab 2814 33± 7

Cistus albidus 0.27± 0.02 a 18.51± 1.86 d 10008± 1011 ab 2476 26± 5

Cistus monspeliensis 0.25± 0.02 a 5.95± 0.43 ab 11342± 832 abc 3904 49± 7

Viburnum tinus 0.21± 0.01 a 45.09± 10.01 h 12448± 797 bcd 1781 134± 4

Arbutus unedo 0.19± 0.02 a 31.48± 1.31 f 14655± 1080 cde 3582 97± 16

Pinus canariensis 0.71± 0.12 b 1.54± 0.20 a 16212± 2099 de 8398 130± 7

Quercus coccifera 0.15± 0.01 a 25.65± 3.60 e 17015± 1102 e 2465 84± 2

Summer

Cistus albidus 0.37± 0.04 bcd 15.86± 2.85 cd 7129± 618 a 1515 69± 5

Cistus monspeliensis 0.39± 0.03 cd 4.85± 0.31 ab 7595± 585 a 2867 63± 2

Ceratonia siliqua 0.31± 0.02 abc 26.89± 1.56 e 8602± 427 ab 1480 108± 3

Pistacia lentiscus 0.29± 0.01 abc 10.51± 0.56 bc 9049± 236 ab 1001 89± 3

Pinus pinea 0.63± 0.04 f 1.25± 0.03 a 9133± 622 ab 1525 166± 3

Viburnum tinus 0.29± 0.03 abc 53.48± 9.79 f 9166± 842 ab 2063 82± 5

Cistus crispus 0.30± 0.04 abc 7.50± 1.40 ab 9243± 1092 ab 2675 50± 2

Cistus salviifolius 0.26± 0.03 abc 10.98± 1.13 bc 10457± 1328 abc 3253 73± 8

Quercus suber 0.23± 0.01 ab 26.90± 1.72 e 11298± 442 bc 2168 75± 3

Arbutus unedo 0.23± 0.01 ab 20.49± 2.83 de 11566± 559 bc 1938 109± 4

Pinus pinaster 0.49± 0.02 de 0.94± 0.02 a 11707 ±469 bc 1150 122± 8

Quercus coccifera 0.21± 0.02 a 27.09± 3.54 e 12986± 1380 c 3379 61± 2

Pinus canariensis 0.59± 0.07 ef 1.24± 0.12 a 16593 ±1669 d 7082 134± 3

Cistus albidus 0.94± 0.04 f 14.69± 0.69 c 2759± 118 a 289 20± 3



10680

Fall

Ceratonia siliqua 0.44± 0.04 cde 23.47± 1.34 d 6163± 423 ab 1465 94± 2

Pistacia lentiscus 0.38± 0.03 bcd 9.09± 0.50 bc 7368± 474 b 2010 82± 4

Cistus monspeliensis 0.38± 0.03 bcd 5.12± 0.48 ab 8509± 894 bc 4381 73± 6

Arbutus unedo 0.34± 0.04 abcd 22.95± 2.04 d 8685± 1013 bc 3509 78± 7

Viburnum tinus 0.31± 0.04 abc 41.81± 8.84 e 8871± 1155 bc 2828 107± 6

Pinus pinea 0.61± 0.05 e 1.09± 0.13 a 9852± 834 bc 2044 145± 4

Quercus suber 0.30± 0.04 abc 21.48± 1.69 d 10494± 738 bc 3614 48± 5

Pinus pinaster 0.55± 0.02 e 0.93± 0.03 a 10541± 370 bc 906 113± 3

Cistus salviifolius 0.22± 0.02 ab 9.48± 1.73 bc 12101± 1252 cd 3066 18± 2

Cistus crispus 0.22± 0.02 ab 10.22± 1.14 bc 12294± 1004 cd 2459 40± 10

Quercus coccifera 0.17± 0.01 a 24.20± 2.32 d 15737± 1416 de 3467 56± 6

Pinus canariensis 0.51± 0.07 de 1.07± 0.12 a 19221± 1799 e 7632 140± 6

The SVR values varied from 2161 m-1 (Viburnum tinus

collected in winter) to 19221 m-1 (Pinus canariensis collected

in fall). One-way ANOVA and Duncan’s multiple comparison

test showed highly significant statistical difference between

SVR values of the studied species in every season, forming 5

levels of significance in winter, spring and fall and 4 levels in

summer (Table 2). SVR data estimated in winter, spring and

summer revealed no significant statistical difference between

species of the genus Cistus, whereas in fall Cistus albidus was

statistically separated from the other species of the same genus

(Table 2). SVR values for Quercus species were not

significantly different in both winter and summer. In all seasons

except winter, only Pinus canariensis needles exhibited SVR

values significantly different from the other pine needles,

whilst Pinus pinea and P. pinaster needles had significantly

similar SVR values. Pine needles collected in winter showed

SVR values statistically all different one by one. Overall, C. siliqua leaves presented the lowest SVR values in all seasons,

while P. canariensis needles showed the highest overall values

for the year (Table 2).

Seasonal and spatial variation of SVR

SVR values are presented by season and site in Table 3. One-

way ANOVA (95% confidence level) showed that within each

examined species, there were a significant effect of the season

on SVR values (Table 4). The greatest seasonal variations

affected A. unedo, Q. coccifera, V. tinus and Cistus species. For

A. unedo, Q. coccifera and V. tinus, SVR values varied

seasonally respectively from 3001 to 15036 m-1, from 2997 to

17015 m-1 and from 2161 to 12448 m-1 (Table 3). For C. crispus,

SVR values changed from 2550 to 12294 m-1 and from 3354 to

12101 m-1 for C. salviifolius. Cistus albidus and C. monspeliensis fluctuated respectively from 2178 to 10008 m-1

and from 2900 to 10685 m-1. SVR values that changed the least

throughout the year corresponded to P. lentiscus, P. pinaster

and P. pinea. Indeed, these values changed from 6450 to 9852

m-1 for P. pinea, from 8295 to 11707 m-1 for P. pinaster and

from 4797 to 9640 m-1 for P. lentiscus.

Seasonal SVR change within species occurred according to H

in all seasons or at least in three seasons for all species

examined except P. canariensis (Table 2). Indeed, for all

species except P. canariensis, SVR increased in spring when H

decreased and during the next seasons, SVR changed following

the evolution of H (i.e. SVR decreased when H rose and

increased when H decreased). However, for P. canariensis

needles, SVR mean values rose continuously throughout the

year from winter to fall though H remained substantially

unchanged. For C. siliqua and P. lentiscus, SVR increased in

spring when H decreased. During the following seasons, H

remained quasi-steady and SVR values also remained steady

and close to each other accordingly. For P. pinaster, P. pinea,

Q. coccifera and Q. suber, SVR followed the evolution trend of

H in only three seasons (Table 2).

One-way ANOVA indicated a significant effect of site on SVR

values within species as regards species sampled in several sites

(C. monspelienis, P. canariensis, P. lentiscus and Q. suber) as

shown in Table 5. The greatest spatial variation is recorded for

P. canariensis whose average SVR values changed in winter

from 4945 (S5) to 14004 m-1 (S6). In spring, its average SVR

changed from 5572 to 25598 m-1 in the same respective sites.

In summer, the SVR mean value for P. canariensis was 7757

m-1 at S5 but increased to 21038 m-1 at S2. Likewise, in fall, P. canariensis SVR mean value varied from 9288 at S5 to 24979

m-1 at S2 (Table 3). Cistus monspeliensis, P. lentiscus and Q. suber, present in most sites, showed volatile swings regarding

average SVR values from a site to another but in a lesser extent.

The lowest spatial variation in winter was recorded for C. monspeliensis, moving from 2381 m-1 at S3 to 3355 m-1 at S6.

In spring, Q. suber SVR values changed the least (from 8324 at

S5 to 10583 m-1 at S4), whereas P. lentiscus exhibited the

lowest SVR variation from one site to another in both summer

and fall. SVR values of P. lentiscus moved respectively in

summer and fall from 7993 (S5) to 9669 m-1 (S3) and from

4872 (S3) to 8955 m-1 (S2) (Table 3).



10681

Table 3: Fresh foliar SVR of 13 forest fuels ± standard error given by season and site. SVR mean values are followed by their

corresponding standard deviations (s.d.)

Season Site Species Fresh SVR± s.e. (m-1) s.d.

Winter Larache (S1) Cistus salviifolius 3354± 135 330

Pinus pinea 6450± 156 348

Quercus suber 2583± 491 1099

Ahl Srif (S2) Ceratonia siliqua 2169± 210 469

Cistus crispus 2550± 158 388

Cistus monspeliensis 3095± 197 482

Pinus canariensis 11040± 560 1252

Pistacia lentiscus 4358± 599 1339


Souk L’Qolla (S3) Arbutus unedo 3001± 167 373

Ceratonia siliqua 2846± 281 629

Cistus albidus 2178± 266 595


Pinus pinaster 8295± 285 638


Tanaqoub (S4) Cistus monspeliensis 3186± 381 853


Dardara (S5) Arbutus unedo 3189± 540 1208




Bellota (S6) Cistus monspeliensis 3355± 308 690


Quercus coccifera 2997± 199 444

Viburnum tinus 2161± 314 703

Spring Larache (S1) Cistus salviifolius 10004± 1149 2814

Pinus pinea 8764± 828 2028























10682




Summer Larache (S1) Cistus salviifolius 10457± 1328 3253

Pinus pinea 9133± 622 1525
























Fall Larache (S1) Cistus salviifolius 12101± 1252 3066

Pinus pinea 9852± 834 2044























10683




Table 4: Analyses of variance regarding season effect on SVR

Species Source of variation SS d.f. MS F F test

Arbutus unedo Method 7...19297882221 3 28266.6666297799 130.113 0.000

significant Error 11811124723.462 377 31329243.298

Ceratonia siliqua Method 7686.69777286. 3 2672.67172877 .22988 0.000


Cistus albidus Method 67878267862727 3 2035025368.715 173.021 0.000


Cistus crispus Method 4521759444.260 3 1507253148.087 208.828 0.000


Cistus monspeliensis Method 8264522345.089 3 2754840781.696 208.234 0.000


Cistus salviifolius Method 5662495051.498 3 1887498350.499 113.136 0.000


Pinus canariensis Method 421216521.808 3 140405507.269 22.959 0.000

significant

Error 1675665201.243 274 6115566.428

Pinus pinaster Method 1347449590.798 3 449149863.599 85.645 0.000


Pinus pinea Method 300193524.571 3 100064508.190 6.382 0.000


Pistacia lentiscus Method 1440818084.452 3 480272694.817 40.875 0.000


Quercus coccifera Method 4290721143.824 3 1430240381.275 188.944 82888


Quercus suber Method 2768733435.808 3 922911145.269 124.013 0.000


Viburnum tinus Method 1622691983.395 3 540897327.798 72.345 0.000


Modelling the relationship between SVR and fuel H

Since summer is the season when wildfires are the most

recurrent, the adjustments of SVR for the effect of H were

performed for summer measurements only. The adjustments

were carried out per species and within species for each

statistically different site regarding the relationship SVR-H

according to Scheffé’s multiple range test, at 95% confidence

level. The resulting model functions are presented in Table 6.

The Pearson's correlation test between SVR and leaf and needle

H revealed significant to highly significant correlation for all

species examined with the exception of C. crispus, for which

the correlation of SVR with H was not significant; therefore,

the regression model of SVR vs H for this species is not

included in Table 6.

All p-values of the Pearson’s correlation between SVR and fuel

H were significant at 99% confidence level and changed from

0.74 (C. monspeliensis at S6) to 0.97 (P. lentiscus at S2).

Adjusted R² values were generally greater than 60% (Table 6).

Therefore, model functions shown in Table 6 seem to be the

most suitable for SVR adjustment for the effect of H. Adjusted

R² was chosen as a criterion for differentiation between the

possible regression models for each modelling because it takes

into account the number of variables, in contrast with R².

Globally, the variation of SVR has been on an upward trend

during the progressive drying of leaves and needles (Figure 2).

This variation is reflecting an increase in SVR when the leaves

and needles lose moisture.



10684

Su

rfa

ce a

rea

-to

-vo

lum

e ra

tio

(m

-1)

Su

rfa

ce a

rea

-to

-vo

lum

e ra

tio

(m

-1)

Moisture content (%)

(b) (c)

(e) (f)

(p) (q) (r)

(g)

(j) (k) (l)

(m) (n) (o)

(h) (i)

(a)

(d)



10685

Su

rfa

ce a

rea

-to

-volu

me

ra

tio

(m

-1)

Moisture content (%)

Figure 2: Surface area-to-volume ratio variation (SVR; m-1) according to leaf and needle moisture content (H; %) per species. For

each species, only the sites statistically different from the point of view of the relation SVR – H are displayed.

(a): Quercus suber at S4; (b): Q. suber at S1; (c): Q. suber at S5; (d): Q. coccifera at S6; (e): Cistus albidus at S3; (f): C. monspeliensis at S4; (g): C. monspeliensis at S6; (h): C. salviifolius at S1; (i): Pinus canariensis at S6; (j): P. canariensis at S5; (k): P. pinaster at S3; (l): P. pinea at S1; (m): Ceratonia siliqua at S2; (n): C. siliqua at S3; (o): Pistacia lentiscus at S5; (p): P. lentiscus at S2; (q): Arbutus unedo at S3; (r): A. unedo at S5; (s): Viburnum tinus at S6

S1: Larache; S2: Ahl Srif; S3: Souk L’Qolla; S4: Tanaqoub; S5: Dardara; S6: Bellota

Table 5: Analyses of variance regarding site effect on SVR values for species collected in several sites

Season Species Source of variation SS d.f. MS F F test

Fall Cistus monspeliensis Method 1350246475.570 3 450082158.523 65.845 0.000 significant

Error 4258531048.481 623 6835523.352

Pinus canariensis Method 927245162.733 2 463622581.367 135.019 0.000 significant

Error 714219659.589 208 3433748.363

Pistacia lentiscus Method 1291060039.465 2 645530019.733 100.373 0.000 significant

Error 3022717060.594 470 6431312.895

Quercus suber Method 118579346.989 3 39526448.996 5.219 0.001 significant

Error 4529394912.388 598 7574238.984

Summer Cistus monspeliensis Method 170406428.099 3 56802142.700 9.204 0.000 significant

Error 1598356104.958 259 6171259.093


Error 1210926527.207 129 9387027.343


Error 1119421835.985 195 5740624.800


Error 1202587871.357 250 4810351.485

Winter Cistus monspeliensis Method 1084141122.573 3 361380374.191 26.430 0.000 significant

Error 5578554916.534 408 13672928.717


Error 3884682413.897 348 11162880.500


Error 2818713452.849 324 8699732.879


Error 4111836618.808 444 9260893.286

Spring Cistus monspeliensis Method 1216580188.869 2 608290094.434 44.095 0.000 significant

Error 4579964343.501 332 13795073.324

(s)



10686

Pinus canariensis Method 1659842055.194 1 1659842055.194 332.816

0.000 significant

Error 713178634.640 143 4987263.179


Error 2891508208.677 250 11566032.835


Error 1651296642.332 392 4212491.435

Table 6: Regression models of the surface area-to-volume ratio vs leaf and needle moisture content of some

dominant Mediterranean natural fuels

Species Site* (altitude) Model function Adjusted R² (%) Pearson’s

correlation p-value

Arbutus unedo Dardara S5 (400m) SVR=-74.13H+17220 78 -0.89**

Souk L'Qolla S3 (250m) SVR=-64.88H+18359 91 -0.96**

Ceratonia siliqua Souk L'Qolla S3 (250m) SVR=-49.05H+16147 89 -0.94**

Ahl Srif S2 (140m) SVR=-287H+13629; H<15% 63 -0.80**

SVR=-19.79H+9732; H≥15% 63 -0.81**

Cistus albidus Souk L'Qolla S3 (250m) SVR=-175H+12068; H<5% 54 -0.76**

SVR=-88.47H+11676; H≥5% 86 -0.93**

Cistus monspeliensis Tanaqoub S4 (615m) SVR=-1100H+20462; H≤14% 70 -0.84**

SVR=-5.19H+5112; H>14% 83 -0.93**

Bellota S6 (130m) SVR=-409H+12743; H<14% 53 -0.74**

SVR=-17.16H+7273; H≥14% 82 -0.92**

Cistus salviifolius Larache S1 (25m) SVR=-241H+11062; H≤18% 80 -0.90**

SVR=-11.85H+6888; H>18% 82 -0.91**

Pinus canariensis Dardara S5 (400m) SVR=-371H+17555; H<1% 84 -0.93**

SVR=-71.80H+17290; H≥1% 91 -0.95**

Bellota S6 (130m) SVR=-545H+48208; H≤24% 62 -0.79**

SVR=-87.79H+37150; H>24% 78 -0.88**

Pinus pinaster Souk L'Qolla S3 (250m) SVR=-52.50H+18538 75 -0.87**

Pinus pinea Larache S1 (25m) SVR=-40.31H+15412 75 -0.87**

Pistacia lentiscus Dardara S5 (400m) SVR=-60.51H+14877 82 -0.91**

Ahl Srif S2 (140m) SVR=-87.58H+17523 94 -0.97**

Quercus suber

Tanaqoub S4 (615m) SVR=-1218H+19786; H<5% 67 -0.82**

SVR=-72.39H+14146; H≥5% 75 -0.88**

Larache S1 (25m) SVR=-901H+19036; H<7% 58 -0.76**

SVR=-19.42H+13009; H≥7% 61 -0.79**

Dardara S5 (400m) SVR=-630H+20628; H≤9% 64 -0.81**

SVR=-43.59H+15148; H>9% 66 -0.82**

Quercus coccifera Bellota S6 (130m) SVR=-361H+19156; H<12% 63 -0.81**

SVR=-65.46H+15744; H≥12% 81 -0.90**

Viburnum tinus Bellota S6 (130m) SVR=-39.95H+13793 76 -0.88**

* For each species, only the sites statistically different from the point of view of the relation SVR – H are displayed. **Significant correlation at the level 0.01 SVR: surface area-to-volume ratio of the leaf (m-1); H: leaf moisture content (%)



10687

The observation of the evolution curves of SVR vs H (Figure

2) and the table of regression models (Table 6) led to

distinguish two separate groups:

1) Species characterized by a linear rise in SVR as they dry,

showing a proportional linearly response to the hydric

stress intensity;

2) Species expressing some resistance to the SVR increase

as a response to hydric stress, mostly at the beginning of

drying.

Overall, the regression models of SVR vs H that fitted the most

to the species examined were grouped in one single piece linear

regression models (1st group) and one piecewise linear models

(2nd group) with thresholds changing from 1 to 24% as shown

in Table 6. The linear model described a linear relationship

between SVR and H characterized by a constant increase in

SVR at drying, from the beginning to the end of the

experimental procedure. This group included A. unedo, C. siliqua (S3), P. lentiscus, P. pinaster, P. pinea and V. tinus. The

behaviours of the second group of species, adjusted with

piecewise linear models, were characterized by an evolution in

two steps. A first step at the beginning of drying where SVR

increased slowly, followed by a second step of higher and

quicker increase in SVR, triggered as soon as the leaf and

needle H dropped below a critical threshold (1 – 24%,

depending on the species and the site as shown in Table 6). The

species belonging to this group were C. siliqua (S2), C. albidus,

C. monspeliensis, C. salviifolius, P. canariensis, Q. coccifera

and Q. suber.

DISCUSSION

Physical parameters of some dominant Mediterranean

forest fuels

Arbutus unedo mean leaf thicknesses measured in all seasons

except winter (0.19 mm in spring, 0.23 in summer, 0.34 in fall)

were close to values found by Yadav et al. [25] (0.26 mm). Leaf

thickness of C. albidus and C. monspeliensis samples measured

by Gillon et al. [26] (respectively 0.55 and 0.54 mm) were in

the seasonal variation interval of leaf thickness of that species

(respectively from 0.27 and 0.25 mm to 1.27 and 0.91 mm) as

presented in Table 2. Mean leaf thicknesses for P. lentiscus

collected in spring and summer (0.29 mm in spring and

summer) were very close to values found by Yadav et al. [25]

(0.28 mm). However, the leaves of P. lentiscus collected in fall

showed the same mean thickness as Gratani et al. [27] (0.38

mm). Quercus coccifera average leaf thickness was either

higher or lower than leaf thickness mentioned in Yadav et al.

[25] and Gillon et al. [26] depending on the collection season.

However, Q. coccifera leaf thickness in summer (0.21 mm) was

very close to the findings of Gillon et al. [26] (0.27 mm).

Quercus suber mean leaf thickness measured by Mediavilla et

al. [28] (0.33 mm) was in the seasonal variation interval of this

species as shown in Table 2 and was close to our values found

for fall samples (0.30 mm).

SVR values of each species and each season were compared

with data found in the literature, having been collected in the

same season. The latter data were obtained by different authors

using different experimental procedures for samples collected

in different sites of the Mediterranean rim. Fall samples of A. unedo showed lower SVR values (5983 m-1), at Souk L’Qolla

(S3), than those obtained by Dimitrakopoulos and Panov [6]

(6585 m-1), whereas at Dardara (S5) our SVR values (11387 m-

1) were greater. Regarding the summer samples, our SVR

values were higher than those measured by Hachmi et al. [4] at

the same season. Moreover, except for C. albidus, which

revealed SVR values close to the values presented in the

literature, the other Cistus species (C. crispus, C. monspeliensis

and C. salviifolius) showed higher values at similar sampling

periods of the year [1,4,6]. Autumnal SVR measurements

regarding P. lentiscus (8955 and 8276 m-1) at two sites (Ahl Srif

S2 and Dardara S5) respectively were greater than those found

by Papió and Trabaud [1] (7030 m-1) and Dimitrakopoulos and

Panov [6] (3532 m-1). However, at S3, the average SVR value

(4872 m-1) was close to the last one reported. In summer, the

SVR value for P. lentiscus at S5 (7993 m-1) was similar to that

reported by Hachmi et al. [4] (7544 m-1). However, the SVR

values computed at S2 and S3 were higher (9485 and 9669 m-1

respectively). Regarding the summer leaves of Q. suber, except

for the leaves collected at Tanaqoub (S4), which had SVR

(8945 m-1) close to the values reported by Hachmi et al. [4]

(8887 m-1), SVR values at the three other sites were greater

(11685, 11405, 13158 m-1). Quercus coccifera and V. tinus

samples reached higher SVR values (15737 and 9166 m-1

respectively) than the values presented by Dimitrakopoulos and

Panov [6] (4141 m-1 for Q. coccifera) and Hachmi et al. [4]

(7510 m-1 for V. tinus) in similar collection seasons.

Concerning pines, the value for P. canariensis (16593 m-1) was

approximately close to that of Hachmi et al. [4] (13004 m-1).

Moreover, the SVR of P. canariensis needles exhibited the

greatest SVR values among all pine species examined. This

feature would hint that this P. canariensis was the most

flammable. Although, P. canariensis needles are known to be

more drought-tolerant than the other pines [29], which is

usually a characteristic of fire resistance according to White

and Zipperer [3], by dint of its drought-coping mechanisms

especially the structure and positioning of stomata [30,31].

Nevertheless, generally, pine needles are substantially

flammable once their H drops below 100% [32]. Pine needles

collected for this work in summer were just above that

threshold (166% for P. pinea; 122% P. pinaster; 133% P. canariensis).

Seasonal and spatial variation of SVR

SVR values, which fluctuated the most throughout the year,



10688

decreased substantially in winter. In this season, leaf H is

usually higher according to Pellizzaro et al. [33], who found,

high time to ignition values for C. monspeliensis at the same

season and then low ignitability, which is in line with our

results showing that the lowest SVR values were recorded in

winter. In summer and fall, when leaf H is low [33], SVR

values are high. This period coincides with low time to ignition

values for C. monspeliensis, thereby increasing their

ignitability [33], which is in line with our findings regarding

SVR values of C. monspeliensis in summer and fall, which

were the greatest of the year. A similar behaviour is common

to other species such as C. siliqua, Q. suber and P. canariensis,

but with less fluctuation throughout the year. The greatest SVR

values were recorded in summer and even in fall, because of

the drought that lasts until autumn. According to Pellizzaro et

al. [33], live fine fuel H began to go back up only after the first

autumn rains. Pistacia lentiscus is among the species that

changed the least. The latter outcome is in accordance with the

findings of Pellizzaro et al. [33], which noted that P. lentiscus

had the least variable H and time to ignition values from one

season to the next.

From winter to spring, H-values of Cistus species sharply

dropped as the spring sampling period coincided with a

heatwave, which became recurrent in spring during the last

years in Morocco. This reaction was particularly more visible

for Cistus species, undoubtedly because of their semi-

deciduous character. Indeed, the summer heat usually make

Cistus species lose some of their leaves unlike other species

(evergreen sclerophyllous), which seem to be more resistant to

the heat; i.e. they are able to maintain leaf turgor longer [34].

Globally, SVR followed the evolution trend of H over the

seasons. This is confirmed by the highly significant correlations

observed between SVR and H as shown in Table 6 even though

measurements were not performed on the same samples during

the four seasons. SVR decreased when H rose and increased

when H decreased. This is evident given that within species the

highest SVR values correspond to the thinnest samples [4] and

thus to the driest [18,19].

The leaves sampled in spring, which are commonly young, are

thinner than old ones [35]. This is the reason why the leaf

thickness in spring was the lowest over the year for some

species such as. The presence of thin leaves in spring for some

species, compared to the leaves collected for the same species

in other seasons, is due to the dominance of young leaves

amongst the samples. This is even more likely that the species

leaves cannot last more than one year. The samples of species

whose leaves can last more than 1 year are less likely to contain

young leaves. Indeed, A. unedo has the particularity to maintain

its leaves only for 11 months (the lowest leaf lifespan among

evergreen sclerophyllous species) [36], which means that

winter samples are leaves at full expansion. In spring, the latter

are replaced by young leaves after bud break. The leaves at full

expansion are characterised by a well-developed palisade

parenchyma, relatively high palisade/spongy parenchyma and

palisade parenchyma thickness to mesophyll thickness [36]

characteristic of species adapted to xeric conditions [17,37].

However, in spring (several days after bud break), since the

leaves are still young, this adaptive strategy has not been

implemented yet [36]. This explains the low thickness of spring

leaves once the heatwave took place, expressed in our

procedure by a decrease in H. This singularity of A. unedo

leaves, as compared to other evergreen sclerophyllous species,

takes springs from the fact that this species is in the borderline

between semi-deciduous to drought and sclerophyllous species;

i.e. the mature leaves show the lowest sclerophylly values while

the young ones are deprived of adaptive characteristics [36].

Viburnum tinus leaves has a short lifespan of ~369 days [38].

The spring leaves of V. tinus are therefore all young since the

leaves coming from the bud break of a year ago are all dead. As

a result, spring leaves of this species were the thinnest over the

year. The spring samples of Q. coccifera contained some young

leaves (turning a red color) coming from a recent bud break.

These leaves are thinner than mature ones [39]. This explains

the low thickness values in spring followed by greater values in

summer since the summer samples were more mature. The

young leaves coming from the spring bud break.

SVR mean values of P. canariensis needles showed an upward

trend over the seasons, although the needles H remained

substantially unchanged. Pinus canariensis needles have

different anatomical and physiological specificity than other

pine needles [29–31]. Besides being classified into a leaf

shrinkage category different from other pine needles [40], P. canariensis needles also have a different response to water

stress [29–31,40].

Site effect was significant on SVR for the species collected in

several sites (Table 5). Statistical significance of site effect was

also highlighted during investigations performed by Pausas et

al. [5] about some flammability traits within species collected

in different sites. In our experiment, regarding species sampled

in different sites, the highest or the lowest SVR values did not

always correspond to the same site for all species. This result is

in accordance with Pausas et al. [5], considering other

flammability traits, whose highest or lowest values did not

correspond to the same sites too. Additionally, according to

Hulshof and Swenson [41], leaf traits examined for 10 forest

species showed variation according to the site. Sclerophylly

and leaf structural traits are also site-dependent and varied

according to the environmental factors [18,25]. Indeed, leaf

tissues thickness and specific leaf mass [25], leaf H [1] and

generally leaf anatomical and biochemical traits [34] vary all

by the season and site. The site and season dependence would

also be linked to leaf dimorphism for P. lentiscus [42] and semi-

deciduous species such as Cistus species [25,43]. Other species

characteristic showing site-dependence, the leaf water

potential, which is variable within species from one site to

another and throughout the year from one season to another

[18]. Since SVR is inversely proportional to leaf and needle

thickness [4], the influence of the H on leaf and needle



10689

thickness [18,19], would also affect SVR. Consequently, the

latter would change likewise according to the season and site

within the same species, which is in accordance with our

outcomes.

Even though sites where the samples were collected are quite

close, there are a significant site effect. The distance between

sites adopted in our experiment are comparable to the distance

between sites chosen by Pausas et al. [5]. Moreover, Hulshof

and Swenson [41] highlighted the site effect on leaf traits

within species though the sites were located in the same forest.

Environmental conditions may vary drastically from one site to

another, changing thereby the species characteristics despite

the short distance between sites.

Modelling the relationship SVR – H

The regression models of the relation SVR vs H should reflect

the SVR variation in all possible situations of leaf and needle

H, especially in summer when the leaves and needles may reach

critical levels of H, mostly under the Mediterranean climate

[34,43,44]. These regression models would also feign the SVR

of dead leaves and needles drying on the forest floor (litter),

which is worth investigating, especially as it is known that litter

is the most flammable forest fuel [21]. The behaviour of SVR

vs H is in line with those of Búrquez [17], Bussotti et al. [18]

and Bacelar et al. [19] who specified that the decrease of H

would reduce the thickness of the leaves and needles, thereby

raising their SVR. In other words, since the thickness is

inversely related to SVR [4], the decrease of the leaf and needle

thickness results in the increase of SVR.

Mediterranean plants make use of several mechanisms to

endure drought conditions [34,45]. The distinction between two

groups of species regarding their response to water stress is also

reported by Bussotti et al. [34], who highlighted two different

water use strategies in Mediterranean plants: evergreen-

sclerophylls vs semi-deciduous species.

According to Búrquez [17], there is a direct linear relation

between the H of leaves and needles and their thickness, which

means that the thickness decreases and thus SVR increases with

the decrease in H, referring to the first group. Nevertheless, this

linear relationship may be impaired due to some drought-

coping mechanisms of the plant [17]. In fact, drought-tolerant

features help the plants to withstand the effects of drought and

maintain the leaf and needle turgor as possible, alluding to

second group [18,19,34,45,46].

SVR vs H behaviour in the second group shows a first stage of

resistance in the increase of SVR before a step of more

substantial increase which is in line with the drought tolerance

trait, common in Mediterranean sclerophylls and reported by

Bussotti et al. [34]. Indeed, the drought tolerance is effective in

the short term, but as the plant is drying, it runs out of water

[34]. The slight increase in SVR at the beginning of drying

implies that some resistance mechanisms are in place to

maintain leaf and needle turgor under hydric deficit conditions,

reflecting drought tolerance and evoking a more conservative

water use.

The behaviour of the second group of species may be explained

by the existence of drought-coping strategies which mitigate

the water stress effects. Indeed, the sclerophyllous leaves of

Quercus species with their thick leaf cuticle, are susceptible to

endure the effects of drought [25,46] and limit the reduction of

the thickness due to water stress [18,19]. In addition, species

with large seeds, such as species of the genus Quercus, are

known to be drought-tolerant [47]. Consequently, the leaves of

Q. coccifera and Q. suber mitigate the rise in SVR under water

deficit conditions as shown in Figure 2a,b,c,d.

The presence of trichomes on the leaves of Cistus species helps

them to cope with water stress effects [19,25,34,43], giving

them the resistance to the increase in SVR observed at the

beginning of drying (Figure 2e,f,g,h). Drought tolerance is

therefore noticeable in Cistus species though they are drought-

deciduous and then less sclerophyllous [25] and their drought

tolerance contributes to leaf abscission in summer [43].

However, drought-deciduousness is seen as a favourable

characteristic allowing the plant to endure drought at the

expense of sclerophylls (e.g. A. unedo, C. siliqua, P. lentiscus,

V. tinus; Figure 2m,n,o,p,q,r,s) in long and intense dry periods

[34], which was simulated by the protocol followed in this

work.

SVR corresponding to P. canariensis needles increased slowly

under the effect of drying and rose significantly later, whilst

SVR of P. pinea and P. pinaster needles rose substantially and

quickly, early in the drying process (Figure 2i,j,k,l). These

findings can be attributed to the more significant drought-

resistant character of P. canariensis needles than the other pine

species [29,31] by dint of a combination of characteristics [29–

31]. Indeed, P. canariensis needles are characterized by a

special shape of the epistomatal chamber inhibiting

transpiration [30], a particular stomata morphology besides

being inserted in such a way to restrict water loss [31]. In

addition, P. canariensis needles have low cuticular

transpiration [29].

Ceratonia siliqua leaves had the distinction of belonging to

both groups according to the sampling site (Table 6; Figure

2m,n). Since C. siliqua leaves are evergreen-sclerophyllous,

their sclerophylly might be affected by the environmental

conditions [18]. Therefore, the response to drought in this

species may be expressed more or less depending on the site,

as shown by our findings. Furthermore, although P. lentiscus

leaves are among the less resistant species to SVR increase

(first group), they have greater water use efficiency and lower

photosynthetic imbalance during drought according to Bussotti

et al. [34]. Such features would favour P. lentiscus over

Quercus species in drought conditions [34] despite the high

sclerophylly [25] and the resistance to SVR increase of

Quercus species, as shown in Figure 2o,p. Arbutus unedo

(Figure 2q,r) has been subject to speculation among the authors



10690

who have expressed doubts about its inclusion among the

sclerophyllous [25,36,48]. In our experiment, A. unedo leaves

fell in the category of the first group (species characterized by

a linear rise of SVR in one single piece).

CONCLUSIONS

SVR values changed significantly according to both site and

season. Such variations were also noted in other plant traits

such as sclerophylly, leaf tissues thicknesses and specific leaf

mass, leaf and needle water potential, leaf and needle H and

generally leaf and needle anatomical and biochemical traits.

The site and season dependence emerges through the leaf

dimorphism in some species such as P. lentiscus and semi-

deciduous species (e.g. Cistus spp.). Statistical correlation

between SVR and H was highly significant in all the species

examined except for C. crispus. Regarding the regression

models of the relation SVR vs H, two groups of species were

observed. The first group included species with SVR values

increasing linearly in one single stage according to drying

intensity and the second group composed of species showing

some resistance to SVR increase with a two-stage behaviour.

Indeed, SVR increased slowly at the beginning of the drying

process, indicating a behaviour of drought tolerance, but once

H became critical, the resistance ran out of steam and thus SVR

rose strikingly as the drying.

Future research prospects are conceivable in order to enlarge

the database of SVR vs H regression models by covering all the

Mediterranean tree and shrub species. This database would be

therefore more useful for the improvement of the fire behaviour

prediction systems because it would give a specific answer

appropriate to any perennial species found in the Mediterranean

ecosystems.

ACKNOWLEDGMENTS.

We gratefully thank the students M’barek Alibouch, Mhammed

Halloumi, George Kwasi Arhin and Abderrahim Raji from

Ecole Nationale Forestière d’Ingénieurs (Salé, Morocco) for

their help during the field and laboratory work. We also

acknowledge Mr. Mohammed Rhaz, laboratory technician, for

technical assistance.

REFERENCES

[1] Papió, C., and Trabaud, L., 1990, "Structural

characteristics of fuel components of five

Meditarranean shrubs", For. Ecol. Manage., 35, pp.

249–259. doi:10.1016/0378-1127(90)90006-W.

[2] Fernandes, P. M., and Rego, F. C., 1998, "A New

Method to Estimate Fuel Surface Area-to-Volume

Ratio Using Water Immersion", Int. J. Wildl. Fire, 8(3),

pp. 121–128.

[3] White, R. H., and Zipperer, W.C., 2010, "Testing and

classification of individual plants for fire behaviour:

Plant selection for the wildlandurban interface", Int. J.

Wildl. Fire, 19, pp. 213–227. doi:10.1071/WF07128.

[4] Hachmi, M., Sesbou, A., Benjelloun, H., and

Bouanane, F., 2011, "Alternative equations to estimate

the surface-to-volume ratio of different forest fuel

particles", Int. J. Wildl. Fire. 20, pp. 648–656.

doi:10.1071/WF09042.

[5] Pausas, J. G., Alessio, G. A., Moreira, B., and

Corcobado, G., 2012, "Fires enhance flammability in

Ulex parviflorus", New Phytol., 193, pp. 18–23.

doi:10.1111/J.1469-8137.2011.03945.X.

[6] Dimitrakopoulos, P. I., and Panov, A. P., 2001, "Pyric

properties of some dominant Mediterranean vegetation

species", Int. J. Wildl. Fire., 10, pp. 23–27.

[7] Rothermel, R. C., and Anderson, H. E., 1966, "Fire

spread characteristics determined in the laboratory",

USDA For. Serv. Intermt. For. Range Exp. Stn. Res.

Pap. INT-30. 34 p.

[8] Rundel, P., 1981, "Structural and chemical components

of flammability", Proc. of the Conference on Fire

Regimes and Ecosystems Properties, Honolulu, HI,.

[9] Dimitrakopoulos, A. P., 2001, "A statistical

classification of Mediterranean species based on their

flammability componentes", Int. J. Wildl. Fire., 10, pp.

113–118.

[10] Finney, M. A., 1998, "FARSITE : Fire Area Simulator

— Model Development and Evaluation", USDA For.

Serv. Res. Pap., 47. doi:U.S. Forest Service Research

Paper RMRS-RP-4 Revised.

[11] Andrews, P. L., 2014, "Current status and future needs

of the BehavePlus Fire Modeling System", Int. J.

Wildl. Fire., 23, pp. 21–33.

[12] Linn, R., Reisner, J., Colman, J. J., and Winterkamp, J.,

2002, "Studying wildfire behavior using FIRETEC",

Int. J. Wildl. Fire., 11, 233. doi:10.1071/WF02007.

[13] Finney, M. A., 2003, "Calculation of fire spread rates

across random landscapes", Int. J. Wildl. Fire., 12, pp.

167–174. doi:10.1071/WF03010.

[14] Cruz, M. G., and Fernandes, P. M., 2008,

"Development of fuel models for fire behaviour

prediction in maritime pine (Pinus pinaster Ait.)

stands", Int. J. Wildl. Fire., 17, pp. 194–204.

doi:10.1071/WF07009.

[15] Hernando, C., Guijarro, M., and De Los Santos, J. A.,

1995, "Determinacion de la Relacion

Superficie/Volumen de las Aciculas Muertas", Invest.

Agr.: Sist. Recur. For., 4(1), pp. 73-85.



10691

[16] Pereira, J., Sequeira, N., and Carreiras, J., 1995,

"Structural-Properties and Dimensional Relations of

Some Mediterranean Shrub Fuels", Int. J. Wildl. Fire.,

5(1), pp. 35-42. doi:10.1071/WF9950035.

[17] Búrquez, A., 1987, "Leaf thickness and water deficit in

plants: A tool for field studies", J. Exp. Bot., 38, pp.

109–114. doi:10.1093/jxb/38.1.109.

[18] Bussotti, F., Bettini, D., Grossoni, P., Mansuino, S.,

Nibbi, R., Soda, C., and Tani, C., 2002, "Structural and

functional traits of Quercus ilex in response to water

availability", Environ. Exp. Bot., 47, pp. 11–23.

doi:10.1016/S0098-8472(01)00111-3.

[19] Bacelar, E. A., Correia, C. M., Moutinho-Pereira, J. M.,

Gonçalves, B. C., Lopes, J. I., and Torres-Pereira, J. M.

G., 2004, "Sclerophylly and leaf anatomical traits of

five field-grown olive cultivars growing under drought

conditions", Tree Physiol., 24, pp. 233–239.

doi:10.1093/treephys/24.2.233.

[20] Dimitrakopoulos, A. P., and Papaioannou, K. K., 2001,

"Flammability assessment of Mediterranean forest

fuels", Fire Technol., 37, pp. 143–152.

doi:10.1023/A:1011641601076.

[21] Liodakis, S., Kakardakis, T., Tzortzakou, S., and

Tsapara, V., 2008, "How to measure the particle

ignitability of forest species by TG and LOI",

Thermochim. Acta., 477, pp. 16–20.

doi:10.1016/j.tca.2008.08.003.

[22] Keane, R. E., 2015, Wildland Fuels Fundamentals and

Applications, Springer, doi:10.1007/978-3-319-09015-

3.

[23] Van Wilgen, A. B. W., Higgins, K. B., and Bellstedt,

D. U., 1990, "The Role of Vegetation Structure and

Fuel Chemistry in Excluding Fire From Forest Patches

in the Fire-Prone Fynbos Shrublands of South Africa",

J. Ecol., 78, pp. 210–222.

[24] Eriksson, I., Teketay, D., and Granström, A., 2003,

"Response of plant communities to fire in an Acacia

woodland and a dry Afromontane forest, southern

Ethiopia", For. Ecol. Manage., 177, pp. 39–50.

doi:10.1016/S0378-1127(02)00325-0.

[25] Yadav, R., Bosabalidis, A., and Vokou, D., 2004, "Leaf

structural features of Mediterranean perennial species:

plasticity and life form specificity", J Biol Res., 2, pp.

21–34.

[26] Gillon, D., Joffre, R., and Ibrahima, A., 1994, "Initial

litter properties and decay rate: a microcosm

experiment on Mediterrranean species", Can. J. Bot.,

72, pp. 946–954. doi:10.1139/b94-120.

[27] Gratani, L., Covone, F., and Larcher, W., 2006, "Leaf

plasticity in response to light of three evergreen species

of the Mediterranean maquis", Trees - Struct. Funct.,

20, pp. 549–558. doi:10.1007/s00468-006-0070-6.

[28] Mediavilla, S., Escudero, A., and Heilmeier, H., 2001,

"Internal leaf anatomy and photosynthetic resource-use

efficiency: interspecific and intraspecific

comparisons", Tree Physiol., 21, pp. 251–259.

doi:10.1093/treephys/21.4.251.

[29] Pardos, M., Calama, R., Climent, J., 2009, "Difference

in cuticular transpiration and sclerophylly in juvenile

and adult pine needles relates to the species-specific

rates of development", Trees, 23,pp. 501–508.

[30] Zellnig, G., Peters, J., Jiménez, M. S., Morales, D.,

Grill, D., and Perktold, A., 2002, "Three-dimensional

reconstruction of the stomatal complex in Pinus canariensis needles using serial sections", Plant Biol.,

4, pp. 70–76. doi:10.1055/s-2002-20438.

[31] Grill, D., Tausz, M., Pöllinger, U., Jiménez, M. S., and

Morales, D., 2004, "Effects of drought on needle

anatomy of Pinus canariensis", Flora, 199, pp. 85–89.

[32] Alexander, M. E., and Cruz, M. G., 2012, "Assessing

the effect of foliar moisture on the spread rate of crown

fires", Int. J. Wildl. Fire., 22, pp. 415–427.

doi:10.1071/WF12008-CO.

[33] Pellizzaro, G., Duce, P., Ventura, A., and Zara, P.,

2007, "Seasonal variations of live moisture content and

ignitability in shrubs of the Mediterranean Basin", Int.

J. Wildl. Fire., 16, pp. 633–641.

doi:10.1071/WF05088.

[34] Bussotti, F., Ferrini, F., Pollastrini, M., and Fini, A.,

2014, "The challenge of Mediterranean sclerophyllous

vegetation under climate change : From acclimation to

adaptation", Environ. Exp. Bot., 103, pp. 80–98.

doi:10.1016/j.envexpbot.2013.09.013.

[35] Bergen, J. Y., 1904, "Relative transpiration of old and

new leaves of the Myrtus type", Bot. Gaz., 38, pp. 446–

451.

[36] Gratani, L., and Ghia, E., 2002, "Changes in

morphological and physiological traits during leaf

expansion of Arbutus unedo", Environ. Exp. Bot., 48,

pp. 51–60. doi:10.1016/S0098-8472(02)00010-2.

[37] Christodoulakis, K., and Mitrakos, N., 1987, Structural

analysis of eleven evergreen phanerophytes in Greece,

In: W. C. Tenhunen et al., eds., Plant Response to

Stress Funct. Anal. Mediterr. Ecosyst., NATO ASI S,

Springer, Berlin, pp. 547–551.

[38] Navas, E., Ducout, M. L., Roumet, B., Richarte, C.,

Garnier, J., and Garnier, J., 2003, "Leaf life span,

dynamics and construction cost of species from

Mediterranean old-fields differing in successional



10692

status", New Phytol., 159, pp. 213–228.

doi:10.1046/j.1469-8137.2003.00790.x.

[39] Manetas, Y., Petropoulou, Y., Psaras, G. K., and

Drinia, A., 2003, "Exposed red (anthocyanic) leaves of

Quercus coccifera display shade characteristics",

Funct. Plant Biol., 30, pp. 265–270.

doi:10.1071/FP02226.

[40] Essaghi, S., Hachmi, M., Yessef, M., and Dehhaoui,

M., 2016, "Leaf shrinkage : a predictive indicator of the

potential variation of the surface area ‑ to ‑ volume

ratio according to the leaf moisture content",

Springerplus, 5(1). doi:10.1186/s40064-016-2900-3.

[41] Hulshof, C. M., and Swenson, N. G., 2010, "Variation

in leaf functional trait values within and across

individuals and species: An example from a Costa

Rican dry forest", Funct. Ecol., 24, pp. 217–223.

doi:10.1111/j.1365-2435.2009.01614.x.

[42] Barazani, O., Dudai, N., and Golan-Goldhirsh, A.,

2003, "Comparison of mediterranean Pistacia lentiscus

genotypes by random amplified polymorphic DNA,

chemical, and morphological analyses", J. Chem.

Ecol., 29, pp. 1939–1952.

doi:10.1023/A:1024862614345.

[43] Werner, C., Correia, O., and Beyschlag, W., 1999,

"Two different strategies of Mediterranean macchia

plants to avoid photoinhibitory damage by excessive

radiation levels during summer drought", Acta

Oecologica, 20, pp. 15–23. doi:10.1016/S1146-

609X(99)80011-3.

[44] Castro, F. X., Tudela, A., and Sebastià, M. T., 2003,

"Modeling moisture content in shrubs to predict fire

risk in Catalonia (Spain)", Agric. For. Meteorol., 116,

pp. 49–59. doi:10.1016/S0168-1923(02)00248-4.

[45] Chaves, J., Maroco, M. M., and Pereira, J. P., 2003,

"Understanding plant responses to drought — from

genes to whole plant", Func Plant Biol Review :, Funct.

Plant Biol., 30, pp. 239–264. doi:10.1071/FP02076.

[46] Aranda, I., Ramírez-Valiente, J. A., and Rodríguez-

Calcerrada, J., 2014, "Características funcionales que

influyen en la respuesta a la sequía de las especies del

género Quercus: variación inter- e intra-específica",

Ecossistemas, 23, pp. 27–36.

[47] Sánchez-Gómez, D., Valladares, F., and Zavala, M. A.,

2006, "Performance of seedlings of Mediterranean

woody species under experimental gradients of

irradiance and water availability: Trade-offs and

evidence for niche differentiation", New Phytol., 170,

pp. 795–806. doi:10.1111/j.1469-8137.2006.01711.x.

[48] Gratani, L., Ghia, E., 2002, "Adaptive strategy at the

leaf level of Arbutus unedo L. to cope with

Mediterranean climate", Flora - Morphol. Distrib.

Funct. Ecol. Plants, 197, pp. 275–284.

doi:10.1078/0367-2530-00041.

Documents

Spatial and Seasonal Response of the Surface Area-to ... · Rabat-Instituts, 10000 Rabat, Morocco. 4 . Département de Statistique et Informatique Appliquées, Institut Agronomique